
TikTok has introduced a new feature that allows politicians and public figures to detect AI-generated videos in which their image is used without permission. The initiative aims to curb the spread of disinformation, especially ahead of elections in Europe, where AI-manipulated videos have become a concern for democratic processes.
The feature, called “Likeness Detection,” began being activated this Friday and is aimed at content creators, politicians, government members and political party accounts. Through it, public figures can receive notifications when videos or accounts suspected of using their AI-generated likeness appear on TikTok. To use the service, they must first register and verify their identity through a trusted external partner.
If they find that their image has been used without consent, they can report the content directly to TikTok’s security teams, which will review it and decide whether to intervene. The initiative comes as Europe faces disinformation campaigns supported by artificial intelligence, including deepfake videos, which can show politicians making statements or taking actions that did not happen. Over the next 12 months, nine other European countries are expected to hold elections.
The European Union began enforcing, on August 2, transparency rules for artificial intelligence, which require deepfake videos and other AI-generated content to be identified as such. TikTok has also long required users to label realistic photos, videos and audio recordings produced with artificial intelligence.
A second tool, free even for those who do not have TikTok
In addition to the feature for public figures, the company has also introduced “Identify AI-Generated Content,” a service that allows citizens to check whether a photo or video was created or altered with the artificial intelligence tools of TikTok and its parent company ByteDance.
The tool can be used for free, even without an account on the platform. It checks material produced with apps such as CapCut, Dreamina, Hypic and Pippit, looking for digital data and invisible markers placed in AI-generated content.
Users can upload photos up to 20 MB and videos or audio recordings up to 100 MB. According to TikTok, the material is deleted immediately after the check and is not used to train artificial intelligence models.
However, the company warns that this service cannot detect every fake. It identifies only content created or modified with TikTok and ByteDance technologies, while material produced with other programs may escape the check.
